Evaluation of tooth angulation measured on cone beam … · 2014-10-16 · Evaluation of tooth angulation measured on cone beam computed tomography and panoramic radiographs A thesis

Evaluation of tooth angulation measured on

cone beam computed tomography and

panoramic radiographs

A thesis submitted in partial fulfilment of the requirements for the degree of

Doctor of Clinical Dentistry

by

Dr Ed Karim

BDS, MFDSRCSI, B.Sc.Dent(Hons)

Orthodontics

School of Dentistry

Faculty of Health Sciences

The University of Adelaide

2014

1

Table of Contents

List of figures ..................................................................................................... 3

List of tables ....................................................................................................... 6

Acknowledgments ............................................................................................. 7

Signed Statement............................................................................................... 8

Summary ........................................................................................................... 9

Introduction ..................................................................................................... 10

Literature review ............................................................................................. 11

Importance of root parallelism ............................................................................................ 11

Panoramic radiography ........................................................................................................ 14

The focal trough (Image layer, layer of sharpness) .............................................................. 17

The relation between focal troughs and dental arches ....................................................... 18

Cone beam computed tomography ..................................................................................... 20

Panoramic radiography and CBCT interface ........................................................................ 21

Aims ................................................................................................................. 22

Hypothesis ............................................................................................................................ 22

Materials and Methods .................................................................................... 23

Method to establish Aim 1: Establish that CBCT is an effective way to assess tooth

angulations ........................................................................................................................... 23

Statistical methods ........................................................................................................................ 28

Method to establish Aim 2: Define a method to establish brand specific OPG focal trough.

.............................................................................................................................................. 28

Method to establish Aim 3: Generate a method to reveal how dentists mis-interpret the

OPG image and to produce OPG and Focal Trough Specific angulations extrapolation. .... 31

Method to establish Aim 4: Compare the angulations extracted from the OPG

extrapolation to the Focal Trough Specific extrapolation. .................................................. 39

Statistical methods ........................................................................................................................ 40

Method to establish Aim 5: Correct the OPG interpretation by forming an OPG brand-

specific formula. ................................................................................................................... 40

Results ............................................................................................................. 41

Results for Aim 1: Establish that CBCT is an effective way to assess tooth angulations ..... 41

2

Statistical analysis ......................................................................................................................... 43

Results for Aim 2: Define a method to establish the brand specific OPG focal trough ....... 44

Results for Aim 3: Generate a method to reveal how dentists mis-interpret the OPG and to

produce OPG and Focal Trough Specific angulations extrapolations. ................................. 45

Results for Aim 4: Compare the angulations extracted from the OPG extrapolation to the

Focal Trough Specific extrapolation. .................................................................................... 49

Visual comparison ......................................................................................................................... 49

Angular measurement comparison .............................................................................................. 51

Statistical analysis ......................................................................................................................... 54

Results for Aim 5: Correct the OPG interpretation by forming an OPG brand-specific

formula. ................................................................................................................................ 59

Discussion ........................................................................................................ 61

Discussion for Aim 1: Establish that CBCT is an effective way to assess tooth angulations 61

Discussion for Aim 2: Define a method to establish the brand specific OPG focal trough . 61

Discussion for Aim 3: Generate a method to reveal how dentists mis-interpret the OPG

and to produce OPG and Focal Trough Specific angulations extrapolations. ...................... 61

Discussion for Aim 4: Compare the angulations extracted from the OPG extrapolation to

the Focal Trough Specific extrapolation............................................................................... 64

Visual comparison ......................................................................................................................... 64

Angle measurements comparison ................................................................................................ 64

Discussion for Aim 5: Correct the OPG interpretation by forming an OPG brand-specific

formula. ................................................................................................................................ 65

Carestream .................................................................................................................................... 65

Vatech ........................................................................................................................................... 66

Conclusion ....................................................................................................... 68

Conclusion for Aim 1: Establish that CBCT is a good way to assess teeth angulations ....... 68

Conclusion for Aim 2: Define a method to establish the brand specific OPG focal trough . 68

Conclusion for Aim 3: Generate a method to reveal how dentists mis-interpret the OPG

and to produce OPG and Focal Trough Specific extrapolations. ......................................... 68

Conclusion for Aim 4 and Aim 5: Compare the angulations extracted from OPG

extrapolation to the Focal Trough Specific extrapolations and Correct the OPG

interpretation by forming an OPG brand-specific formula. ................................................. 68

References ....................................................................................................... 70

3

List of figures

Figure 1 long axes relation in a “normal” untreated individual from "Six keys to normal

occlusion".[1] ....................................................................................................................................... 11

Figure 2 A flat curve of Spee requires parallel roots.[1] .................................................................... 12

Figure 3 The inter-related effect of tip and torque[1]. ....................................................................... 13

Figure 4 Burstone’s explanation of how tooth inclination is indicative of a problem which is dental

(left) or skeletal (right) in origin[6]. .................................................................................................... 14

Figure 5 Rotational tomography from Patero[15]. ............................................................................. 15

Figure 6 Orthopantomography as described by Paatero. O1 and O3 represent the axes for the right

and left buccal segments while O2 represents the axis for the narrower anterior segment. X is the x-

ray beam. R is the x-ray source. K is the film and 1,2,3,4 represent the film movement sequence.[17]

.............................................................................................................................................................. 16

Figure 7 Different focal troughs of 3 different machines as illustrated by Lund and Manson-Hing

[29] ........................................................................................................................................................ 18

Figure 8 OPG of skull/typodont used in the study by Mckee et al [46]. ............................................. 19

Figure 9 Method validation proposal for Aim 1 ................................................................................ 23

Figure 10 Teeth with titanium cylindrical (2mm height and diameter) markers inserted at the

coronal and apical ends. Red lines are drawn between coronal and apical markers on the four sides

of each tooth. ........................................................................................................................................ 24

Figure 11 Aim 1 method validation block formation. ........................................................................ 25

Figure 12 Teeth in wax block. ............................................................................................................. 25

Figure 13 The block inside the grid box. ............................................................................................ 26

Figure 14 The block was photographed from two orthogonal angles. .............................................. 26

Figure 15 Using the Cobb angle measurement tool in RadiAnt DICOM Viewer. ............................ 27

Figure 16 Focal trough grid orientation in the OPG machine. ........................................................ 29

Figure 17 The focal trough detection block on a dental arch grid. ................................................... 30

Figure 18 OPG of the focal trough detection block. Note that some rods are blurred or out of focus

while a particular band is in focus. ..................................................................................................... 30

Figure 19 Marking of the most medial and the most lateral rods in focus on the 2.5mm grid. ....... 31

Figure 20 (A) OPG in the mind of dentists (B) Red plane represent the long axis of the upper left

permanent canine. ................................................................................................................................ 32

4

Figure 21 The skull after cutting the alveolar processes. ................................................................... 33

Figure 22 Stainless steel balls inserted at the coronal and apical ends of extracted teeth. ............... 34

Figure 23 The skull with inserted wax typodont. .................................................................................. 34

Figure 24 Skull OPGs (A)Vatech machine (B) Carestream machine. .............................................. 35

Figure 25 Using Dolphin® Imaging 11.5 Premium Marking the stainless steel with red dots and

obtaining their X,Y,Z coordinates. ...................................................................................................... 36

Figure 26 (A) Frontal view of teeth long axes in green. Coronal and apical ends in red.(B)

Superior view of teeth within the skull.(C)Vatech focal trough and(D) CareStream focal trough

planes wrapped around the teeth coordinates. .................................................................................... 37

Figure 27 Matching spline curve to focal trough plane and projecting root coordinates to the curve.

.............................................................................................................................................................. 38

Figure 28 intersecting the curve with the coordinates orthogonal projections. ................................ 39

Figure 29. The focal troughs (green shading) and central planes (red line) for the Carestream® and

Vatech® machines. The grid measure is 2.5mm x 2.5mm. ................................................................... 44

Figure 30. (A) Carestream OPG extrapolation. (B) Vatech OPG extrapolation. ................................ 45

Figure 31. (A) Carestream Focal Trough Specific extrapolation. (B) Vatech Focal Trough Specific

extrapolation. ........................................................................................................................................ 46

Figure 32 OPG extrapolations visual comparison for Carestream and Vatech. ............................... 49

Figure 33 Focal Trough Specific extrapolations visual comparison for Carestream and Vatech. .. 50

Figure 34 Carestream OPG & Focal Trough Specific extrapolations visual comparison. .............. 50

Figure 35 Vatech OPG & Focal Trough Specific extrapolations visual comparison. ...................... 51

Figure 36 Agreement between Carestream Focal Trough Specific extrapolation and OPG

extrapolation. The horizontal scale is the number of the angles from Table 8. The vertical scale is

the angle measurements. ...................................................................................................................... 53

Figure 37 Agreement between Vatech Focal Trough Specific and OPG extrapolations. The

horizontal scale is the number of the angles from Table 8. The vertical scale is the angle

measurements. ...................................................................................................................................... 53

Figure 38 Agreement between Carestream and Vatech OPG extrapolations. The horizontal scale is

the number of the angles from Table 8. The vertical scale is the angle measurements. ................... 54

Figure 39 Correlation agreement Carestream and Vatech Focal Trough Specific extrapolations.

The horizontal scale is the number of the angles from Table 8. The vertical scale is the angle

measurements. ...................................................................................................................................... 54

5

Figure 40 Focal Trough Specific extrapolation comparison between Carestream and Vatech

measured in angles. .............................................................................................................................. 55

Figure 41 OPG extrapolation comparison between Carestream and Vatech measured in angles. .. 55

Figure 42 Comparison between Carestream Focal Trough Specific and OPG extrapolations

measured in angles. .............................................................................................................................. 56

Figure 43 Comparison between Vatech Focal Trough Specific and OPG extrapolations measured

in angles. .............................................................................................................................................. 56

Figure 44. Skull OPGs (A) Vatech machine (B) Carestream machine. Note that the Green arrows

point at the medial side of the mandibular condylar head and the blue arrows point at the lateral side

of the mandibular condylar head. Also note how the same condylar head look different in two

different panoramic machine images. ................................................................................................... 63

Figure 45 Dental arches (D) and mandibular bone arches (M) plotted by gender (A) and by

ethnicity (B) from Nummikoski et al.[73] ........................................................................................... 67

6

List of tables

Table 1. Photo measurements in angles from T1, T2 and T3 to the three dimensions X (height), Y

(width) and Z (depth). 1, 4 & 6 represent the central incisor, premolar and molar, respectively. In

columns, for example, F Tooth-X means the angle between the tooth and X axis as seen from the

Frontal view of the wax block inside the grid box. R Tooth-Z means the angle between the tooth and

Z axis looking from the Right side view of the wax block inside the grid box. ................................... 41

Table 2 Averages from the measurements In Table 1. .......................................................................... 42

Table 3. CBCT measurements in angles from T1, T2 and T3. 1, 4 & 6 represent the central incisor,

premolar and molar respectively. In columns, for example, F Tooth-X means the angle between the

tooth and X axis looking from the Posterior view in the CBCT machine. R Tooth-Z means the angle

between the tooth and Z axis looking from the Right side view in the CBCT machine. ...................... 42

Table 4. Averages from the measurements in Table 3. ......................................................................... 42

Table 5. Least Squares Means for the photo and CBCT measurements. .............................................. 43

Table 6. Least Squares Means differences for the photo and CBCT measurements. ........................... 43

Table 7. Coordinates values for coronal and apical markers of teeth. For example; UR7C means

Upper Right second molar crown. UR6R Upper Right first molar Root. ............................................. 47

Table 8. Teeth inclinations and angles nomination as measured on OPG and Focal Trough Specific

extrapolations for both Carestream and Vatech machines. ................................................................... 52

Table 9. Limits of agreement and P values for 4 comparisons of Carestream and Vatech Focal Trough

Specific and OPG extrapolations. ......................................................................................................... 57

Table 10. Group Least Square Means for OPG extrapolations of Carestream and Vatech machines. . 58

Table 11. Group Least Square Means for Focal Trough Specific extrapolations of Carestream and

Vatech machines. .................................................................................................................................. 58

Table 12. Group Least Square Means for Focal Trough Specific and OPG extrapolations of

Carestream machine. ............................................................................................................................. 58

Table 13. Group Least Square Means for Focal Trough Specific and OPG extrapolations of Vatech

machine. ................................................................................................................................................ 59

Table 14. Angle differences between Focal Trough Specific and OPG extrapolations for Carestream

machine. ................................................................................................................................................ 59

Table 15. Angle differences between Focal Trough Specific and OPG extrapolations for Vatech

machine. ................................................................................................................................................ 60

7

Acknowledgments

I wish to express my appreciation and thanks to my supervisors: Professor Wayne Sampson,

Associate Professor Craig Dreyer and Professor Lindsay Richards, for their expert advice,

encouragement and editorial opinion throughout this project.

I would also like to extend my appreciation to the following people and organizations whom

without this project would not have reached its results.

• Dr Michael J Rielly for his generous donation of dry human skull.

• Dr Joseph Moussa for opening his doors for me and making his practice and

equipment available for this study.

• Professor Maciej Henneberg / Prof Anatomical Sciences - The University of Adelaide

for his forensic identification insight.

• The Adelaide Dental Hospital for their technical support.

• Mr Don Chorley for his radiographic expertise input.

• Dr Stephen Langford and InCiDental Imaging for donating their time and equipments.

• Dr Paul Buchholz for donating his time and practice equipment.

• Mr Scott Vallance for his software insight.

• Ms Suzanne Edwards and DMAC, University of Adelaide for their statistical

assistance.

• Dr Balya Sriram and Dr Eugene Twigge for their unconditional help.

• Mrs Lucy Hatch for her endless kindness and willingness to reach a hand whenever

she was asked for help.

• Mr Eddie Sziller and Mr Jon Cor-Udy for laboratory technical assistance.

• My colleagues Dr Berna Kim and Dr Vandana Katyal for their friendship and

collegiality during my time as a post-graduate student.

• My dear friends, near and far who have provided continuous encouragement and

support during the last three years.

Finally, this thesis is dedicated to my mother Dr Suha Al-Farhan not just for being my

mother who brought me to this world but also for being my closest friend, mentor and life

role model showing me how hard work always pays off and even the sky is not the limit.

Thanks Mum.

This project and the degree attached to it and all what perceived it of achievements in my

life only exist because I had my wife, Dr Zainab Hamudi in my life. She is my motive to

reach higher levels and my inspiration to keep developing my professional career. She is

the kind face I return to after a hard day at work to keep reminding me what was all that

about. I love you Zainab and I always will.

I am blessed to have all these people in my life and I thank my God for their existence.

9

Summary

Panoramic radiographs have been used for decades in the field of dentistry and in

orthodontics in particular. Panoramic radiography use in orthodontics includes (but is not

limited to): pre-treatment dental assessment, dental age estimation, the detection of dental

anomalies, the identification of missing and impacted teeth, the prognosis for unerupted teeth,

periodontal tissue assessment and pre-finishing root parallelism. The determination and

validity of root position has been debated by many authors and practitioners. Knowledge

regarding the production of panoramic radiographs questions whether they are the best way

of judging tooth angulation.

Aims: The main aim of this study was to assess dental angulation measurements generated by

two panoramic x-ray machines. An attempt would be made to correct appraisal differences

and, therefore, render the films more useful in providing the necessary diagnostic information

to achieve optimal orthodontic treatment results.

Methods: A dry human skull with an inserted typodont was imaged using two different

panoramic machines in addition to a CBCT machine. Teeth within the typodont had metal

markers attached to their coronal and apical ends. The markers and therefore the long axes of

the teeth were identified. Teeth were divided into 5 groups (Anterior-Anterior, Anterior-

Premolar, Premolar-Premolar, Premolar-Molar and Molar-Molar). The angles between

adjacent teeth were measured on an “OPG extrapolation” produced by tracing markers

directly on panoramic radiographs. The same measurements were performed on “Focal

Trough Specific extrapolations” which were produced by applying the panoramic machine-

specific focal trough around the coordinates of the coronal and apical markers. Each machine

had its own “OPG extrapolation” and “Focal Trough Specific extrapolation”. The four

extrapolations were compared.

Results: A wide range of variation in tooth angulation was found between the measurements

of each machine. For the Carestream machine, the Premolar-Molar and Molar-Molar regions

were represented poorly on panoramic radiographs. While for the Vatech machine, Premolar-

Premolar and Anterior-Premolar were the groups represented least accurately on the

panoramic radiographs.

Conclusion: Panoramic radiographs should be interpreted with caution when assessing tooth

angulations. Panoramic x-ray machine manufacturers should be encouraged to provide a

range of error values to help practitioners have a better understanding of the limitations of

panoramic radiographic interpretation.

10

Introduction

Patient satisfaction and wellbeing are the main goals for health providers and are a major

consideration in treatment planning. A limitation of skills and scientific knowledge may put

clinicians under pressure to take responsibility for identifying an optimal treatment plan for

patients. Treatment options are provided and selection of a suitable treatment plan is

ultimately the patient‟s prerogative. This implies that the patient will trust the clinician during

treatment and in managing the process of treatment. It is an orthodontic imperative to

perform tasks to the limit of current knowledge. These tasks include the proper attachment of

brackets of appropriate slot size, the selection of the optimal wire configuration and the

considered placement of bends to deliver appropriate biomechanical orthodontic forces. The

clinician must ultimately make the final decision regarding the achievement of treatment

goals when treatment may be discontinued and band/brackets removed.

A major objective of orthodontic treatment is to produce ideal tooth inclinations and

angulations within their respective alveolar bone processes and in relation to adjacent teeth.

Several mechanisms have been suggested to assess root positions which include a clinical

examination of crown orientation and sulcus palpation of the final occlusion. A common

method used to determine the mesio-distal root angulation is the use of a pre-finishing

panoramic radiograph to judge whether tooth roots parallel each other.

Andrews in his highly cited paper “The six keys to normal occlusion”[1] suggested the

concept that if a clinician knew what constituted “right”, he could then directly, consistently,

and methodically identify and quantify what was “wrong”. Knowing what is normal and

acceptable allows the determination of what is unacceptable.

11

Literature review

Importance of root parallelism

Andrews[1] indicated that tipping of the long axis of the crown, not the long axis of entire

tooth is the second most important occlusal key after molar relationship. It was indicated that

each untreated, normal occlusion documented had a distal inclination of the gingival portion

of each crown which was considered a constant. The inclination varied with each tooth

type, and within each type, but the tip pattern was consistent between individuals. Andrews

highlighted the need for the long axes of teeth to be tipped distally but did not indicate that

the teeth should be parallel (Figure 1). However, it was stressed that a flat occlusal plane

implied root parallelism (Figure 2).

In the inter-relationship between tip and torque, Andrews stated that for every 4° of lingual

root torque, a 1° of mesial root tip was produced which implied that the effect of root

inclination had an effect in another dimension apart from mesio-distal tipping (Figure 3).

Figure 1 long axes relation in a “normal” untreated individual from "Six keys to normal occlusion".[1]

12

Figure 2 A flat curve of Spee requires parallel roots.[1]

13

Figure 3 The inter-related effect of tip and torque[1].

The American Board of Orthodontics considers root angulation as a criterion for the

assessment of cases submitted for examinations [2]. Although panoramic radiographs are

used for assessment, a negative score would not be ascribed to a canine displaying a lack of

parallelism with adjacent tooth roots when a final panoramic radiographic assessment was

made. This came from a recognition and acknowledgement of the distortion that frequently

occurs within panoramic radiographs. The common mistakes in root angulation occurred in

relation to the maxillary lateral incisors, canines, second premolars, and mandibular first

premolars[3].

Orthodontists have stressed the importance of assessing the parallelism of roots prior to

finishing orthodontic treatment and its importance in preventing relapse in extraction cases.

In addition, the orthodontist should also check for excessive tooth tipping. Graber supported

the use of periodic panoramic radiographs during treatment to achieve optimal treatment

objectives and root position [4].

14

Ursi et al. stated that “a properly treated case, viewed radiographically, should display the

same root arrangement after treatment as a patient presenting with a normal occlusion. The

upper central and lateral incisor roots should be slightly convergent, and the remaining upper

teeth should show a distal inclination, except for the second molars, which should be mesially

tilted. The lower incisors should be upright, and the other lower teeth should be increasingly

distally inclined as one moves posteriorly”[5]. Presumably, Ursi et al were only referring to

root angulation but the concept is mildly confusing.

Burstone [6] stressed the importance of tooth inclination in the differentiation of asymmetric

deviations of skeletal or dental origin. It was indicated that an assessment of the inclination of

teeth to the occlusal plane (as a reference), and by comparing left and right sides, a

discrepancy may be judged as dental (if the teeth have a different relation to the occlusal

plane) or skeletal (same relation to the occlusal plane) component.

Figure 4 Burstone’s explanation of how tooth inclination is indicative of a problem which is dental (left) or skeletal

(right) in origin[6].

Panoramic radiography

The panoramic radiograph is of significant benefit to a dentist. The film provides a

mechanism for identifying dental impactions [7], dental anomalies [8-11] and assists in dental

age estimation [12]. The advantages of using panoramic radiography are identified as its low

radiation dose, low operator time usage, relatively short patient exposure time, and excellent

patient comfort. The disadvantage is that panoramic radiography has limitations related to

magnification and distortion errors particularly associated with the reliability and accuracy in

the assessment of form, size and location of structures within the dental arch [13].

In 1949, Paatero [14] proposed a new method of radiography which evolved to contemporary

panoramic radiography. The concept of tomography was that, in order to capture a round

object (the maxillofacial or cranial structures) radiographically, there needed to be rotational

axes in which either the object, the film or the source stay stationary and the other two rotate.

15

Rotational tomography occurs when a patient is seated in a dental chair and rotates

simultaneously with a suitable curved film, each on its own axis, but in opposite

directions Figure 5 [15]. The procedure is called „pantomography‟ and is a combination of

panoramic and tomographic views). Paatero [14] used either concentric, eccentric or double

eccentric rotational axes to gain the required image.

Figure 5 Rotational tomography from Patero[15].

In 1957, Hudson, Kumpula and Dickson [16] showed that it was possible to keep an object

stationary while controlling the rotation of the x-ray source and the film. This was used for

screening large numbers of personnel in the United States Armed Forces.

Paatero [17] realised that the right and left buccal segments have a wider diameter when

compared with the narrower anterior segment. The technique was revised and combined into

Orthopantomography (OPG) in which there are several rotational axes (Figure 6). The

Panorex unit is known for two centres of rotation and the Orthopantomograph has three

centres of rotation.

16

Figure 6 Orthopantomography as described by Paatero. O1 and O3 represent the axes for the right and left buccal

segments while O2 represents the axis for the narrower anterior segment. X is the x-ray beam. R is the x-ray source. K is

the film and 1,2,3,4 represent the film movement sequence.[17]

Because of the three centres of rotation in Orthopantomographs, there is greater

magnification and less sharpness in the incisor region. Positioning the patient‟s head is

critical and, since the radiation source must penetrate the vertebral column, the variable

density between the incisor and the first premolar area sometimes results in a lack of

uniformity of the image. [13]

Precise assessment of an arch length deficiency and room for eruption in the canine and first

premolar region is not reliably determined and consistent. Varying arch form and

circumference may cause distortion [4]. The Orthopantomograph and the Panorex create an

impression of less available space than actually exists. Individual tooth malposition enhances

this impression. The distortion and magnification that exist in both machines will not change

diagnostic decisions, provided that all diagnostic information is correlated first [4].

The concept of tomography revolves around the x-ray source and the film (receptor) moving

around the object in two different directions and capturing the layer of interest within the

17

object sharply. The layers above and below are blurred [18]. The coordinated movement and

speed of the source and receptor allow the same structure to be captured at the same spot of

the film during radiograph capture while unwanted structures will be blurred. The difference

in x-ray beam angulation accounts for different widths in the focal trough [18].

When anatomical structures are viewed in a panoramic radiograph, the obliquity of the

projection means that the view is not directly at the buccal aspect but with an oblique

angulation of -7° to -9° depending on the setting of the panoramic machine [19]. This could

make a palatally-impacted premolar look like it is mesially impacted.

Scarfe et al and Tronje et al [20, 21] found that different panoramic machines have different

angulations between the focal trough central plane and the central ray of the x-ray beam and

did not assume an optimal 90 degrees angle. Many machines present favourable angulation in

the anterior region but vary from 15 to 45 degrees in the posterior region.

Lucchesi et al [22] indicated that, in assessing inclined objects, plane radiographs are more

accurate than panoramic radiographs.

Garcia [23] and Garcia-Figueroa et al [24, 25] concluded that a change in the buccolingual

angulation of teeth can affect the mesiodistal angulation between adjacent teeth in panoramic

radiographs. Most of the effect was noted in the canine-premolar region and did not affect the

anterior region. Lee [26] confirmed the issue by saying that “if roots appear incorrectly

aligned on panoramic images, the teeth in question should be carefully examined clinically

(or with periapical images) to differentiate the need for appliance adjustments that alter root

tip versus root torque”.

The focal trough (Image layer, layer of sharpness)

A three-dimensional region of acceptable image un-sharpness is unique in width and shape to

each panoramic system and determined by the following factors [27, 28]:

Scanning geometry and effective projection radius.

Source-to-film distance.

Film transport speed.

Focal spot size.

Width of the X-ray beam.

Image receptor.

Average dental arch forms and shapes have been suggested. Lund and Manson-Hing [29, 30]

demonstrated how different machines produce different shapes and dimensions of focal

troughs (Figure 7) and how each would have vertical and horizontal magnification

distortions.

18

Figure 7 Different focal troughs of 3 different machines as illustrated by Lund and Manson-Hing [29]

Rejebian pointed to the magnitude of both horizontal and vertical tooth dimension distortions

on the panoramic radiographs and stated “The panoramic radiograph can be useful to the

practitioner as a clinical aid in the diagnosis of mixed-dentition and space problems, provided

that there is awareness of the relative magnitude of distortion associated with the particular

equipment being employed and that a standardized technique is used”. In his paper, Rejebian

compared the linear records on panoramic radiographs, plaster models and (for vertical

dimension) the lengths of extracted teeth and found that the distortion magnification could

range horizontally between -4% (19% in the article) to 55% and vertically between 23% -

32% [31].

The relation between focal troughs and dental arches

Many authors have tried to determine the average shape of the human dental arch (Hellman

[32], Meredith and Higley[33], Bonwill[34], Hawley[35], Williams[36], Hudson, Kumpula,

and Dickson[16], Lasher[37], Paatero[38], Bromwell[39], Angle[40], Wheeler[41],

19

Black[42], Izard[43,44]. All had diverse concepts of arch form indicating that it was

impossible to compile a single mathematical description of the human dental arch. Patients

attend for treatment with a wide range of variable dental arches. It therefore becomes

questionable to image all patients from one focal trough perspective.

Lund and Manson-Hing and Manson-Hing et al [44, 45] showed that different races, genders

and ages can still fit reasonably well inside the differently-shaped focal troughs of various

panoramic machines.

Ursi et al [5] recommended the orthopantomogram as a valuable tool to help in deciding

whether orthodontic treatment objectives were met. Three different machines were compared

and it was found that different axial inclinations were displayed without causing a clinical

issue. Their references were a line passing through the lowest part of the orbit and a line

passing through the two mental foramina.

Mckee et al [46] in a similar study, compared the outcome of 4 different panoramic machines

to directly measure typodont teeth inserted into a dry skull. The long axes of the teeth were

determined by inserting chromium steel balls into the centre of the most occlusal and the

most apical (centre of bifurcation in multi-rooted teeth) part of the teeth. The reference was a

0.020” stainless steel arch wire ligated to 0.022” slot clear orthodontic brackets bonded to the

teeth (Figure 8).

Figure 8 OPG of skull/typodont used in the study by Mckee et al [46].

Mckee‟s findings were compared with measurements taken by a coordinate measuring

machine (CMM). It was found that the largest root angular difference in the maxilla was

between the canine and the first premolar where their relative root parallelism or convergence

could be projected as root divergence resulting in excessive root convergence if treated to

panoramic radiograph parallelism. In the mandible, a similar but lesser discrepancy was

20

found between the canine and the first premolar. However, these adjacent teeth were the

most highly underestimated of mandibular teeth in relation to their true angulations. The

largest angular difference between adjacent teeth occurred in the mandible between the lateral

incisor and the canine, with relative root parallelism seen on the panoramic radiograph as root

convergence. Treating to the panoramic radiograph in this region could lead to excessive root

divergence. The conclusions match those of McDavid et al [47] and Philipp and Hurst [48]

who found root angles were mostly distorted in the canine and premolar region of both

arches. Mckee et al [46] also determined that, despite the use of different panoramic

machines, with different focal-trough dimensions and beam-projection angles, the machines

appeared to systematically overestimate or underestimate true angulations in a similar

fashion.

Cone beam computed tomography

Van Elslande et al [49] found, by comparing root angulation determined by a coordinate

measuring machine (CMM) and CBCT generated panoramic-like image, that a CBCT can

produce deviated measurements in isolated regions. This could be explained by the way the

references were set (an arch wire) and the curve set for the CBCT scan as the focal trough

following that reference.

Tong et al [50] established the University of South Carolina root vector analysis program in

which metal ball indicators were implanted on coronal and apical ends of teeth set in a

typodont inserted into a dry skull. The long axes of teeth were measured to a reference arch

wire to determine mesio-distal and facio-lingual tooth inclination using a coordinate

measuring machine (CMM). A comparison was then conducted with a CBCT of the same

skull rendered in Dolphin imaging software. It was shown that using their algorithm, CBCT

measurements were accurate and reliable relative to the CMM golden standard. This method

was previously established to measure root angulation and inclination in patients nearing the

end orthodontic treatment [51].

DeHaan et al [52] used CBCT to measure tooth angulations and inclinations to the occlusal

plane. Panmekiate et al [53] found that linear measurements varied between 0.03 to 0.28mm

when 2 different milliamperage and 4 peak kilovoltage settings (8 settings) were compared. It

was advised that using low peak kilovoltage and milliampere values for linear measurements

in the posterior mandible was beneficial.

Taking into account the previous statement, and by applying triangle mathematical rules [54],

if one object is 20 mm in length (a tooth) then it will take 0.35mm of error determining its

distal end (either to the right or left sides) to cause a change of 1° in its angulation. This

means CBCTs range of error falls within the human range of error of locating any specific

point in space.

21

Panoramic radiography and CBCT interface

Wise et al [55] suggested that although panoramic radiographs reconstructed using cone

beam computed tomography (CBCT) volumetric data were better than conventional

panoramic radiographs in assessing mesiodistal root angulations, they were still unreliable

methods for such an assessment.

Pittayapat et al [56] indicated that CBCT scans were more reliable and had a higher level of

agreement when retrospectively compared with pre-treatment panoramic radiographs of 38

patients.

Bouwens et al [57] compared the mesio-distal root angulation of 35 patients following

orthodontic treatment. Panoramic radiographs and CBCT scans were assessed using the

occlusal plane as a reference line and statistically significant differences between the two

methods were found which led the authors to advise caution when interpreting panoramic

radiographs.

Leuzinger et al [58] reviewed the panoramic radiograph of 22 patients who were near the end

of their orthodontic treatment and found that OPGs overestimated the root contacts when

compared with CBCT evaluations.

It is imperative to determine whether panoramic radiographs are a reliable representation of

tooth long axis alignment and, if not, how far they differ from reality. To investigate this

issue, an understanding of what a panoramic radiograph represents is required. A panoramic

film, is often interpreted as a plane wrapped around the face of the patient on a curve

representing the motion of the radiographic machine and rendered as a flat 2-dimensional

plane, which we will call “OPG mis-interpretation”. The present study differs and aims to

reference the mesio-distal inclination of teeth measured to the motion curve (see Aim 3 for

more explanation).

22

Aims

The aims of this study are:

Aim 1: Establish that CBCT is an effective way to assess tooth angulations.

Aim 2: Define a method to establish a brand-specific OPG focal trough.

Aim 3: Generate a method to reveal how dentists mis-interpret an OPG image and to

produce OPG and Focal Trough Specific angulation extrapolations.

Aim 4: Compare the angulations extracted from the OPG extrapolation to the Focal

Trough Specific extrapolation.

Aim 5: Correct the OPG interpretation by forming an OPG brand-specific formula.

Hypothesis

The null hypothesis to be investigated is that panoramic radiographs are a reliable method to

measure tooth angulation relative to each other.

23

Materials and Methods

The project advanced only after each method was established and its related aim was

fulfilled.

Method to establish Aim 1: Establish that CBCT is an effective way to assess

tooth angulations

The purpose of this project was to test whether a CBCT is an acceptable and reliable method

to measure tooth angulations in space. The design for this method is shown in Figure 9

Figure 9 Method validation proposal for Aim 1

The validation of tooth angulations was assessed by the following:

i) Three previously extracted human teeth (lower central incisor, upper premolar and

upper molar) were selected to cover the range of anterior, premolar and molar

teeth.

ii) The teeth were marked with 2mm diameter and 2mm length titanium markers on

their coronal and apical ends to help establish their long axis. The coronal end was

determined by the centre of the incisal edge or the centre of the central fossa. The

24

apical end was determined as the apical tip for a single rooted tooth or the centre

of the furcation area for a multi-rooted tooth.

iii) The coronal and apical markers were connected with red lines on the labial,

mesial, palatal/lingual and distal surfaces of the teeth as shown in Figure 10.

Figure 10 Teeth with titanium cylindrical (2mm height and diameter) markers inserted at the coronal and apical ends.

Red lines are drawn between coronal and apical markers on the four sides of each tooth.

iv) A clear plastic box formed a template to create a validation block. Three titanium

laser welded rods were set in the corner of the box to establish 90° reference

planes in all three dimensions X (height),Y (width) and Z (depth). A block of red

sheet wax was made as in Figure 11 A, B & C.

25

Figure 11 Aim 1 method validation block formation.

v) The three teeth were placed with arbitrary angulation into the block as in Figure

12 A and B.

Figure 12 Teeth in wax block.

vi) A box was lined with 5mm grid sheets to create a standardised measurement

environment.

vii) The titanium rods were aligned precisely with the grid lines from all aspects.

(Figure 13).

26

Figure 13 The block inside the grid box.

viii) The block was photographed from two orthogonal angles by lining the central

axis of the camera lens with the Z axis of the scene (Figure 14).

Figure 14 The block was photographed from two orthogonal angles.

27

ix) Tooth angulations were measured using the following method: As the metal

titanium rods were aligned precisely with grid lines in the background, a

translucent protractor was used to measure the angles between the metal rods

(representing reference axes X, Y and Z) and the red lines on the teeth

(representing their long axis).

x) Those measurements were tested by the same operator (E.K.) three times with no

less than 24 hours between each trial. The results can be found in Table 1 and

Table 2.

xi) The Block was then imaged in a CBCT machine. The CBCT images were viewed

in RadiAnt© DICOM Viewer Medixant Evaluation Version 1.1.8.4646 (64 bit).

The long axes of the teeth were determined by marking the coordinates of the

apical and coronal titanium markers for each tooth and then measured to the

coordinates of the reference X, Y or Z titanium rod using the Cobb angle (which

was originally used to measure coronal plane deformity on antero-posterior plane

radiographs in the classification of scoliosis. The angle was named after the

American orthopaedic surgeon John Robert Cobb (1903–1967) [59, 60] who

devised the measurement tool. This tool allowed measurement of the angle

between four points, 2 on the tooth long axis and 2 on the metal rod) in the CBCT

viewer (Figure 15).

Figure 15 Using the Cobb angle measurement tool in RadiAnt DICOM Viewer.

28

xii) Those measurements were tested by the same operator (E.K.) three times with no

less than 24 hours between each trial. The results can be found in Table 3 and

Table 4.

The results of Aim 1 methodology are presented in the results section.

Statistical methods

The statistical software used was SAS 9.3 (SAS Institute Inc., Cary, NC, USA). An Intra-

class Correlation Coefficient (ICC) was calculated. A linear mixed-effects model was also

created. The least square means were also calculated.

Method to establish Aim 2: Define a method to establish brand specific OPG

focal trough.

Two different OPG machines were used in this study; Carestream® CS 9000 and Vatech®

PAX-Reve3D. Both companies failed to provide a clear definition of their machine‟s focal

troughs; therefore, a method for determining the focal trough needed to be established for

each OPG machine.

A 2.5mm grid with a drawing of an average dental arch form was placed in each OPG

machine and oriented at the level of a patient‟s occlusal plane. This arch form was only taken

as a guide to follow during this experiment. Any other form could have been used and would

have led to the same results. A point where the patient‟s midline bisecting the arch form was

oriented exactly where a patient would bite edge-to-edge anteriorly on the OPG machine‟s

mouth piece when taking an OPG (Figure 16).

29

Figure 16 Focal trough grid orientation in the OPG machine.

Nine pairs of metal rods of 2mm diameter and lengths of 13, 15, 17, 19, 21, 23, 25, 27, 29mm

were inserted 3mm into a wax block in a straight line along the length of the block. The

longest rod pair was placed in the middle and then in length descending order the others were

inserted towards the sides forming one line of rods (Figure 17).

30

Figure 17 The focal trough detection block on a dental arch grid.

The wax block was placed at the most distal point of the grid and perpendicular to the grid

midline (Figure 17A). An OPG was taken (Figure 18) and the most medial and lateral

defined rods on the radiograph marked on the grid (Figure 19). The wax block was moved

two grid units anteriorly (without changing its perpendicular orientation to the midline) and

another OPG was taken and the process repeated until the entire arch was covered. On the

testing OPGs the most focused band was recorded. The arch form was only used as an

approximate guide, for example, in Figure 17B, the shortest rods on the left were in focus not

those in the middle and tallest.

Figure 18 OPG of the focal trough detection block. Note that some rods are blurred or out of focus while a particular

band is in focus.

31

Figure 19 Marking of the most medial and the most lateral rods in focus on the 2.5mm grid.

The areas between the grid markings were highlighted and mirror imaged over the midlines

which identified the focal troughs for the Carestream® and Vatech® machines. The median

distance between the grid markings was calculated to produce a continuous line which

represented the central plane of the focal trough Figure 29.

Method to establish Aim 3: Generate a method to reveal how dentists mis-

interpret the OPG image and to produce OPG and Focal Trough Specific

angulations extrapolation.

Conceptually, an OPG represents an orthogonal view of the dental and maxillofacial complex

spread over a 2D plane that was wrapped on a curve around the head of a patient (Figure

20A).

In reality, the plane is not a simple curve but is a consequence of focal trough geometry.

32

Figure 20 (A) OPG in the mind of dentists (B) Red plane represent the long axis of the upper left permanent canine.

A dry skull was donated to this study by the kindness of Dr Michael J Reilly. Professor

Maciej Henneberg (Prof Anatomical Sciences-The University of Adelaide) estimated the

skull belonged to a male in his third decade of life and was probably of Australo-melanesian

ancestry.

The alveolar processes of the maxilla and mandible were removed (Figure 21). Most of the

hard palate was preserved.

33

Figure 21 The skull after cutting the alveolar processes.

A full set of sound extracted upper and lower permanent teeth (excluding third molars) was

collected. Stainless steel balls of 2mm diameter were imbedded into each tooth at its corornal

end (mid-incisal edge for anterior teeth, buccal or mesiobuccal cusp tip for posterior teeth)

and apical end (apex of root for anterior teeth or mesiobuccal root for posterior teeth) (Figure

22). Stainless steel was found easier to identify on radiographs and CBCT images and caused

less beam scattering than the earlier-used titanium. Previous studies also used coronal and

apical stainless steel markers [22, 24, 25, 46, 49, 50, 61-63].

34

Figure 22 Stainless steel balls inserted at the coronal and apical ends of extracted teeth.

A wax rim was inserted to substitute and represent the removed alveolar processes. The

extracted teeth were set-up in the wax typodont in class I incisal, canine and molar

relationship with upper and lower dental midlines aligned with the mid-sagittal plane of the

skull. The wax typodont allowed for easy dental arch changes. The vertical dimension of the

skull was recorded prior to the removal of the bony alveolar processes and was maintained

after inserting the wax typodont. The final skull is shown in Figure 23.

Figure 23 The skull with inserted wax typodont.

35

OPG images of the skull were obtained by using the two different machines, Vatech® PAX-

Reve3D and Carestream® CS 9000 (Figure 24).

Figure 24 Skull OPGs (A)Vatech machine (B) Carestream machine.

The skull was also imaged with a cone beam computed tomography machine Vatech® PAX-

Reve3D and processed through Dolphin® Imaging 11.5 premium software using the

University of Southern California (USC) root vector analysis program in the Dolphin 3D

36

module (Figure 25). The Dolphin software allowed automatic selection of the centre of the

stainless steel balls.

Figure 25 Using Dolphin® Imaging 11.5 Premium Marking the stainless steel with red dots and obtaining their X,Y,Z

coordinates.

All the markers representing X, Y, Z coordinates were produced by assigning the 0, 0, 0

origin point below the contact points between the upper central incisors. This approximates

the position a patient would occlude on a jig situated in the OPG machine. The relevant co-

ordinates can be found in Table 7.

The following was then created:

OPG extrapolations: A direct tracing of the coronal and apical metal markers on the

panoramic radiographs. One was created for each panoramic machine brand image. They can

be seen in Figure 30.

Focal Trough Specific extrapolations: A 2D representation for the projection of the 3D

coronal and apical metal marker‟s coordinates onto a curve representing the machine specific

focal trough. The extrapolation represents how teeth would line up and relate to each other if

viewed orthogonally through the focal trough. One was created for each panoramic machine

brand focal trough. They can be seen in Figure 31.

This takes into account the effect of the rotation centres during image capture. The OPG is

not created from a simple curved plane with a consistent relation to the dental arches. The

focal trough represents the volume in which the image plane is created.

37

A 3D depiction was created when the coordinates were arranged in space and the respective

specific OPG machine‟s focal trough planes were wrapped just outside the coordinates

(Figure 26). Applying this method assured that these coordinates were viewed within the

Focal Trough Specific extrapolations orthogonally in the same way a dentist would interpret

a panoramic radiograph. The interpretation was limited to and followed the machine-specific

focal trough curve around the coordinates.

Figure 26 (A) Frontal view of teeth long axes in green. Coronal and apical ends in red.(B) Superior view of teeth within

the skull.(C)Vatech focal trough and(D) CareStream focal trough planes wrapped around the teeth coordinates.

To extrapolate the coronal and apical coordinates of teeth on the 3D curve representing the

focal trough wrapped around these coordinates, the coordinates need to be projected onto that

curve horizontally (in the Y axis direction). The extrapolation needed to proceed along a line

intersecting the tooth point and the OPG curve in a direction commensurate with the OPG

curve. If more than one intersection was possible, the closest intersection was taken. Solving

this problem required a number of steps summarised as:

ImageJ Software [64] was used to digitise the points on scanned OPG curve images. An open

source software package (SciPy) [65] was used to fit a spline curve around the coordinates.

38

To transform the 3D to 2D data we simply had to consider the X and Z dimension data and

subtract the Y dimension which would project the locations onto the X, Z plane where the

spline curve (representing the central plane of the focal trough) is defined as in Figure 27.

For this purpose, coordinates of the metal markers were connected (projected) onto the curve

and the shortest connection perpendicular to the curve was considered. The intersection of

that connection with curve was marked as in Figure 28. The curve was then flattened to a 2D

plane to produce the “Focal Trough Specific extrapolation” which can be seen in Figure 31.

Figure 27 Matching spline curve to focal trough plane and projecting root coordinates to the curve.

39

Figure 28 intersecting the curve with the coordinates orthogonal projections.

OPG extrapolations were produced by marking (tracing) the apical and coronal stainless steel

balls directly on the panoramic radiographs.

The CBCT-derived co-ordinates were used to produce the Focal Trough Specific

extrapolations.

Method to establish Aim 4: Compare the angulations extracted from the OPG

extrapolation to the Focal Trough Specific extrapolation.

The proposed comparison was to:

1- Compare the Carestream Focal Trough Specific extrapolation with the Carestream

OPG extrapolation.

2- Compare the Vatech Focal Trough Specific extrapolation with the Vatech OPG

extrapolation.

3- Compare the Carestream OPG extrapolation with the Vatech OPG extrapolation.

4- Compare the Carestream Focal Trough Specific extrapolation with the Vatech Focal

Trough Specific extrapolation.

The comparisons were performed visually by using the “best fit” method. Angle

measurements were also compared by plotting directly on charts. Those comparisons will be

shown in the results section.

40

Statistical methods

The statistical software used was SAS 9.3 (SAS Institute Inc., Cary, NC, USA).

A Bland-Altman analysis [66-68] was used to assess the level of agreement between OPG

and Focal Trough Specific extrapolations for the two brands, Carestream and Vatech. A 95%

limit of agreement was defined as mean difference ±1.96 standard deviations, producing a

lower level and an upper level of agreement, which are shown as red dashed lines in Figure

40-Figure 43 (as is mean difference). The blue solid lines are the clinically relevant cut-offs

for acceptable agreement i.e. an angle difference of plus or minus 5 degrees.

A paired t-test, Wilcoxon Signed Rank test, linear regression test and Least Square means

difference test were performed. The results can be found in the results section.

Method to establish Aim 5: Correct the OPG interpretation by forming an OPG

brand-specific formula.

The formula or the range of error was calculated by establishing the range of difference

between the measurements from the OPG extrapolations to the measurements from the Focal

Trough Specific extrapolations. This will be explained in the results, discussion and

conclusion sections.

41

Results

The results section will be presented, in the same manner as the methodology, following the

aims of the study.

Results for Aim 1: Establish that CBCT is an effective way to assess tooth

angulations

The photo measurements from Aim 1 methodology are presented in Table 1. The Averages of

the photo measurements are presented in Table 2. The CBCT measurements from Aim 1

methodology are presented in Table 3. The averages of those measurements are found in

Table 4.

Table 1. Photo measurements in angles from T1, T2 and T3 to the three dimensions X

(height), Y (width) and Z (depth). 1, 4 & 6 represent the central incisor, premolar and molar,

respectively. In columns, for example, F Tooth-X means the angle between the tooth and X

axis as seen from the Frontal view of the wax block inside the grid box. R Tooth-Z means the

angle between the tooth and Z axis looking from the Right side view of the wax block inside

the grid box.

Photo Trial 1

F Tooth-X (°) F Tooth-Y(°) R Tooth-X(°) R Tooth-Z(°)

1 3.5 86.5 14 76

4 5 85 5 85

6 1 89 1 89

Photo Trial 2

F Tooth-X(°) F Tooth-Y(°) R Tooth-X(°) R Tooth-Z(°)

1 3 87 14.5 75.5

4 6 84 4 86

6 1.5 88.5 2 88

Photo Trial 3


1 2 88 13 77

4 6 84 4 86

6 2 88 2 88

42

Table 2 Averages from the measurements In Table 1.

Photo Average


1 2.8 87.1 13.8 76.1

4 5.6 84.3 4.3 85.6

6 1.5 88.5 1.6 88.3

Table 3. CBCT measurements in angles from T1, T2 and T3. 1, 4 & 6 represent the central

incisor, premolar and molar respectively. In columns, for example, F Tooth-X means the

angle between the tooth and X axis looking from the Posterior view in the CBCT machine. R

Tooth-Z means the angle between the tooth and Z axis looking from the Right side view in

the CBCT machine.

CBCT Trial 1

Tooth F Tooth-X(°) F Tooth-Y(°) R Tooth-X(°) R Tooth-Z(°)

1 5 85 13 77

4 3.5 86.5 5 85

6 3 87 5 85

CBCT

Trial 2


1 5.5 84.5 12.5 77.5

4 3.5 86.5 6 84

6 4 86 6 84

CBCT

Trial 3


1 6 84 12 78

4 4 86 5.5 84.5

6 3 87 5 85

Table 4. Averages from the measurements in Table 3.

CBCT Average


1 5.5 84.5 12.5 77.5

4 3.6 86.3 5.5 84.5

6 3.3 86.6 5.3 84.6

43

Statistical analysis

The statistical analysis was performed by SAS 9.3 software (SAS Institute Inc., Cary, NC,

USA).

An Intra-class Correlation Coefficient (ICC) was calculated to be 0.999 showing strong

reliability between the two methods in measuring angles of the teeth.

A linear mixed-effects model was also created with angle measurement as the outcome

variable, method (photo measurements vs CBCT measurements) as the predictor, and random

effects: tooth ID and tooth ID*view*axis (to account for clustering). Least Square Means and

Least Square Means differences are given below in Table 5 and

Table 6. An example of interpretation is, from the Least Squares Means table: photo

measurements had a mean angle measurement of 45.25 degrees whilst CBCT measurements

had a mean angle measurement of 45.33 degrees. From Difference of Least Squares Means:

photo measurements had a mean angle measurement 0.08 degrees less than CBCT

measurements (mean estimate = -0.8, 95% Confidence Interval: -1.5, 1.3).

Table 5. Least Squares Means for the photo and CBCT measurements.

Least Squares Means

Effect Method Estimate Standard

Error

DF t Value Pr > |t| Alpha Lower Upper

Method Photo 45.2500 12.0018 11 3.77 0.0031 0.05 18.8341 71.6659

Method CBCT 45.3333 12.0018 11 3.78 0.0031 0.05 18.9175 71.7492

Table 6. Least Squares Means differences for the photo and CBCT measurements.

Differences of Least Squares Means

Effect Method Method Estimate Standard

Error

DF t Value Pr > |t| Alpha Lower Upper

Method Photo CBCT -0.08333 0.6211 11 -0.13 0.8957 0.05 -1.4504 1.2837

44

Results for Aim 2: Define a method to establish the brand specific OPG focal

trough

After following the methodology mentioned for aim 2 in the materials and methods section,

a brand specific focal trough representation was achieved for the Carestream and the Vatech

machines. The two different focal troughs varied in shape and dimensions. The results are

shown in Figure 29.

Figure 29. The focal troughs (green shading) and central planes (red line) for the Carestream® and Vatech® machines.

The grid measure is 2.5mm x 2.5mm.

45

Results for Aim 3: Generate a method to reveal how dentists mis-interpret the

OPG and to produce OPG and Focal Trough Specific angulations

extrapolations.

Figure 30. (A) Carestream OPG extrapolation. (B) Vatech OPG extrapolation.

46

Figure 31. (A) Carestream Focal Trough Specific extrapolation. (B) Vatech Focal Trough Specific extrapolation.

The coordinates for the coronal and apical markers were registered using the University of

Southern California (USC) root vector analysis within Dolphin® Imaging 11.5 premium

software. The origin point was set at the contact point between the upper central incisors

when the patient was biting edge-to-edge on the mouth piece when taking an OPG. Patient‟s

midline and the machine‟s midline were coincident as per OPG machine manufacturer‟s

instruction.

The coordinates were registered in the three planes of space; sagittal plane, coronal plane and

occlusal plane. The positive values for the sagittal plane indicate the coordinate being anterior

to the origin point while posterior coordinate had negative values. The positive values for the

coronal plane indicate the coordinate being left to the origin point while right side coordinate

47

had negative values. The positive values for the occlusal plane indicate the coordinate being

superior to the origin point while inferior coordinate had negative values. The coordinates can

be found in Table 7.

The coordinates were then projected to a 2D plane using ImageJ Software [64] and the open

source software package (SciPy) [65]. The results are shown in Figure 31. This allowed for

the production of “Focal Trough Specific extrapolations” which represent what OPGs would

look like if they were really just planes wrapped around the patient‟s head.

Table 7. Coordinates values for coronal and apical markers of teeth. For example; UR7C

means Upper Right second molar crown. UR6R Upper Right first molar Root.

USC 3D Root Research Landmark Name

Coronal plane(mm) Occlusal plane(mm) Sagittal plane (mm)

UR7C -30.6 4 -36

UR7R -25.9 17.4 -38.3

UR6C -27.5 3.7 -25.5

UR6R -24.6 17.3 -28.5

UR5C -28 3.2 -17.4

UR5R -17.9 17.1 -23.8

UR4C -22.1 3.6 -11.8

UR4R -18.2 20.8 -20.5

UR3C -17.3 2.4 -6

UR3R -13.2 23.7 -17.6

UR2C -11.4 2.9 -2.8

UR2R -8.5 20.1 -13.1

UR1C -3.8 2.7 -0.2

UR1R -2.5 21.2 -10.9

UL1C 4.5 2.2 0.6

UL1R 3.9 22.5 -11.7

UL2C 11.9 2.4 -1.8

UL2R 10.9 20.3 -11.1

UL3C 17.5 2.4 -5.7

UL3R 13.7 24.5 -16.2

UL4C 22.1 2.7 -12.3

UL4R 16.9 20.3 -18.5

UL5C 25.1 2 -18.6

UL5R 18.7 18.8 -25.1

UL6C 27.5 3.1 -25.2

UL6R 24.4 15.3 -28

UL7C 30.6 4.1 -34.8

UL7R 27.9 16.7 -40.2

48

LL7C 27.5 -3.5 -32.2

LL7R 31.5 -19.1 -34.8

LL6C 24.2 -4.5 -21.4

LL6R 28.9 -20.3 -25.2

LL5C 22.3 -4.5 -13.8

LL5R 20.2 -20.6 -16.6

LL4C 19.6 -4.1 -7.2

LL4R 14.9 -22.2 -14.5

LL3C 12.6 -2.3 -3.4

LL3R 13.7 -23.4 -5.1

LL2C 6.4 -2.3 -1.1

LL2R 6.7 -23.8 -3.7

LL1C 1.2 -2 0.2

LL1R 1.2 -20.9 -4.2

LR1C -4.3 -2.4 -0.9

LR1R -3.3 -22 -4

LR2C -9.4 -2.8 -1.9

LR2R -8.4 -24.8 -7.2

LR3C -14.9 -2.7 -5

LR3R -12.4 -23.9 -7.5

LR4C -20 -3.7 -9.2

LR4R -15.6 -18.8 -11.3

LR5C -24.3 -3.5 -15.1

LR5R -20 -23.4 -19.9

LR6C -26.2 -4.5 -22.1

LR6R -27.3 -20.5 -26.3

LR7C -28.6 -4 -33.2

LR7R -31 -17.5 -38.1

49

Results for Aim 4: Compare the angulations extracted from the OPG


Visual comparison

Figure 32 to Figure 35 illustrate the visual cross-comparison between different

extrapolations. This was achieved by overlaying the extrapolations to the best fit.

Figure 32 OPG extrapolations visual comparison for Carestream and Vatech.

50

Figure 33 Focal Trough Specific extrapolations visual comparison for Carestream and Vatech.

Figure 34 Carestream OPG & Focal Trough Specific extrapolations visual comparison.

51

Figure 35 Vatech OPG & Focal Trough Specific extrapolations visual comparison.

Angular measurement comparison

If the angle between the upper right second molar and the upper right first molar is identified

as angle 1 and the angle between the upper right first molar and the upper right second

premolar is identified as angle 2, a progression is created through the dental quadrants

(quadrant 1, 2, 3 then 4) until angle 26 is identified (between lower right first molar and

lower right second molar). Teeth were grouped into 3 regions described as anteriors,

premolars and molars. Angles were nominated according to the relation between these

groups. The comparison of those angles over Carestream or Vatech machines OPG or Focal

Trough Specific extrapolations is presented in Table 8.

52

Table 8. Teeth inclinations and angles nomination as measured on OPG and Focal Trough Specific extrapolations for both Carestream and

Vatech machines.

AngleNo.

Angle between teeth:

Angle nomination Carestream Focal Trough Specific extrapolation

Vatech Focal Trough Specific extrapolation

Carestream OPG extrapolation

Vatech OPG extrapolation

1 17-16 Molar-Molar -12.7 -9.2 3.8 4.8

2 16-15 Premolar-Molar 10.8 30.8 -9.2 -9.4

3 15-14 Premolar-Premolar 17.6 -37.4 13.0 15.6

4 14-13 Premolar-Anterior 2.5 11.8 0.2 -0.4

5 13-12 Anterior-Anterior -27.4 -4.7 -3.1 -3.6

6 12-11 Anterior-Anterior -8.9 3.6 -1.7 -2.1



9 22-23 Anterior-Anterior -13.2 1.5 1.8 2.3

10 23-24 Premolar-Anterior 1.6 13.8 5.4 5.4

11 24-25 Premolar-Premolar 12.8 -40.0 -0.1 0.3

12 25-26 Premolar-Molar 9.1 29.6 -2.4 -1.5

13 26-27 Molar-Molar -25.4 -16.9 -14.5 -13.9

14 37-36 Molar-Molar 5.6 3.3 4.0 3.7

15 36-35 Premolar-Molar -4.9 -11.2 -22.1 -22.3

16 35-34 Premolar-Premolar 2.1 5.9 -3.9 -2.7

17 34-33 Premolar-Anterior -7.2 1.7 8.9 7.3



20 31-41 Anterior-Anterior 3.8 -4.6 0.7 0.5


22 42-43 Anterior-Anterior 4.8 6.4 4.3 4.9

23 43-44 Premolar-Anterior 9.4 2.5 4.7 4.8

24 44-45 Premolar-Premolar -25.4 1.4 -6.5 -8.6

25 45-46 Premolar-Molar 2.4 -16.6 -12.4 -13.2

26 46-47 Molar-Molar -12.2 -6.2 -5.2 -6.9

53

Figure 36 shows the level of agreement between Carestream Focal Trough Specific and OPG

extrapolations.

Figure 37 shows the level of agreement between Vatech Focal Trough Specific and OPG

extrapolations.

Figure 38 shows the level of agreement between Carestream and Vatech OPG extrapolations.

Figure 39 shows the level of agreement between Carestream and Vatech Focal Trough Specific

extrapolations.

Figure 36 Agreement between Carestream Focal Trough Specific extrapolation and OPG extrapolation. The horizontal scale is

the number of the angles from Table 8. The vertical scale is the angle measurements.

Figure 37 Agreement between Vatech Focal Trough Specific and OPG extrapolations. The horizontal scale is the number of

the angles from Table 8. The vertical scale is the angle measurements.

54

Figure 38 Agreement between Carestream and Vatech OPG extrapolations. The horizontal scale is the number of the angles

from Table 8. The vertical scale is the angle measurements.

Figure 39 Correlation agreement Carestream and Vatech Focal Trough Specific extrapolations. The horizontal scale is the

number of the angles from Table 8. The vertical scale is the angle measurements.

Statistical analysis

The statistical software used was SAS 9.3 (SAS Institute Inc., Cary, NC, USA).

A Bland-Altman analysis [66-68] plots can be seen in Figure 40-Figure 43 (as too is mean

difference). The upper and lower levels of agreement are shown as red dashed lines and the

clinically relevant cut-offs for acceptable agreement (angle difference of plus or minus 5 degrees)

as blue solid lines.

55

Difference

-50

-40

-30

-20

-10

0

10

20

30

40

50

60

Average

-25 -20 -15 -10 -5 0 5 10 15 20 25

Difference

-50

-40

-30

-20

-10

0

10

20

30

40

50

60

CBCT Comparison: Brand C versus Brand V

Bland-Altman Plot

Figure 40 Focal Trough Specific extrapolation comparison between Carestream and Vatech measured in angles.

Difference

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

6

Average

-25 -20 -15 -10 -5 0 5 10 15 20

Difference

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

6

OPG comparison: Brand C versus Brand V

Bland-Altman Plot

Figure 41 OPG extrapolation comparison between Carestream and Vatech measured in angles.

56

Difference

-25

-20

-15

-10

-5

0

5

10

15

20

25

Average

-25 -20 -15 -10 -5 0 5 10 15 20

Difference

-25

-20

-15

-10

-5

0

5

10

15

20

25

Brand C comparison: CBCT versus OPG

Bland-Altman Plot

Figure 42 Comparison between Carestream Focal Trough Specific and OPG extrapolations measured in angles.

Difference

-60

-50

-40

-30

-20

-10

0

10

20

30

40

50

Average

-25 -20 -15 -10 -5 0 5 10 15

Difference

-60

-50

-40

-30

-20

-10

0

10

20

30

40

50

Brand V comparison: CBCT versus OPG

Bland-Altman Plot

Figure 43 Comparison between Vatech Focal Trough Specific and OPG extrapolations measured in angles.

57

Table 9. Limits of agreement and P values for 4 comparisons of Carestream and Vatech Focal

Trough Specific and OPG extrapolations.

Variable 1 Variable 2 Mean

difference

Lower 95%

limit of

agreement

Upper 95%

limit of

agreement

Paired t-test

P value

Signed Rank

P value

Carestream

Focal Trough

Specific

extrapolation

Vatech Focal

Trough

Specific

extrapolation

-0.40641 -39.0087 38.1959 0.9170 0.2451

Carestream

OPG

extrapolation

Vatech OPG

extrapolation

0.10162 -1.8300 2.0332 0.6037 0.5609

Carestream

Focal Trough

Specific

extrapolation

Carestream

OPG

extrapolation

-0.79453 -23.2088 21.6197 0.7261 0.9312

Vatech Focal

Trough

Specific

extrapolation

Vatech OPG

extrapolation

-0.28650 -36.0652 35.4922 0.9369 0.6836

When a paired t-test was performed, no statistical significance was found between the means of

each pair. When a Wilcoxon Signed Rank test was performed, there was found to be no

statistically significant difference between the mean ranks of each pair. This is probably because

these tests only compare the means of each group. The mean differences were not large (ranging

from -0.79 to 0.10) unlike the limits of agreement. However, the Bland-Altman plots test the

agreement of the 2 pairs in question by evaluating the spread of the data.

For each comparison, linear regression was performed to test the association between the

difference between the pairs and the tooth-tooth comparison assessed (a 5-category nominal

variable) that can be found in Table 9. A statistically significant difference was found between the

Focal Trough Specific extrapolation and the OPG extrapolation for the Carestream machine (P

value=0.0002) and between the Focal Trough Specific extrapolation and the OPG extrapolation

for the Vatech machine (P value=0.0042).

Angle nominations were tested to show if any region within the dental arch was better represented

(Table 10-Table 13). An example of a Least Square means difference is for the Carestream

regression when comparing Focal Trough Specific extrapolation with the OPG extrapolation a

comparison between Anterior-Anterior (-4.3) and Molar-Premolar (15.9) results in an angle

difference of 20.3 degrees more in the Molar-Premolar comparison (P value<0.0001).

It should be noted that the Anterior-Anterior measure represent the average of all Anterior-

Anterior relationships.

58

Table 10. Group Least Square Means for OPG extrapolations of Carestream and Vatech machines.

Group Least Squares Means

Group Estimate Standard

Error

z Value Pr > |z|

Anterior-Anterior 0.1787 0.2899 0.62 0.5376

Anterior-Premolar 0.5665 0.4584 1.24 0.2166

Premolar-Premolar -0.5095 0.4584 -1.11 0.2664

Premolar-Molar 0.08990 0.4584 0.20 0.8445

Molar-Molar 0.06680 0.4584 0.15 0.8842

Table 11. Group Least Square Means for Focal Trough Specific extrapolations of Carestream and

Vatech machines.



Error

z Value Pr > |z|

Anterior-Anterior -3.3479 5.4899 -0.61 0.5420

Anterior-Premolar -5.9049 8.6803 -0.68 0.4963

Premolar-Premolar 19.3342 8.6803 2.23 0.0259

Premolar-Molar -3.7733 8.6803 -0.43 0.6638

Molar-Molar -3.9278 8.6803 -0.45 0.6509


Carestream machine.



Error

z Value Pr > |z|




Premolar-Molar 15.9727 4.1108 3.89 0.0001

Molar-Molar -8.2396 4.1108 -2.00 0.0450

59


Vatech machine.



Error

z Value Pr > |z|




Premolar-Molar -3.7733 8.6803 -0.43 0.6638

Molar-Molar -3.9278 8.6803 -0.45 0.6509

Results for Aim 5: Correct the OPG interpretation by forming an OPG brand-

specific formula.

The differences between OPG and Focal trough Specific extrapolations were calculated and can

be found for Carestream and Vatech machines in Table 14 and Table 15 respectively.

Table 14. Angle differences between Focal Trough Specific and OPG extrapolations for

Carestream machine.

Angle number


Angle nomination Carestream Focal Trough Specific extrapolation

Carestream OPG extrapolation

Difference

1 17-16 Molar-Molar -12.7 3.8 -16.6

2 16-15 Premolar-Molar 10.8 -9.2 20.1

3 15-14 Premolar-Premolar 17.6 13.0 4.5

4 14-13 Premolar-Anterior 2.5 0.2 2.2

5 13-12 Anterior-Anterior -27.4 -3.1 -24.2


7 11-21 Anterior-Anterior -0.9 -1.9 0.9


9 22-23 Anterior-Anterior -13.2 1.8 -15.0

10 23-24 Premolar-Anterior 1.62 5.4 -3.8

11 24-25 Premolar-Premolar 12.8 -0.1 13.0

12 25-26 Premolar-Molar 9.1 -2.4 11.5

13 26-27 Molar-Molar -25.4 -14.5 -10.8

14 37-36 Molar-Molar 5.6 4.0 1.5

15 36-35 Premolar-Molar -4.9 -22.1 17.2


60

17 34-33 Premolar-Anterior -7.2 8.9 -16.2



20 31-41 Anterior-Anterior 3.8 0.7 3.1




24 44-45 Premolar-Premolar -25.4 -6.5 -18.8

25 45-46 Premolar-Molar 2.4 -12.4 14.9

26 46-47 Molar-Molar -12.2 -5.2 -7.0

Table 15. Angle differences between Focal Trough Specific and OPG extrapolations for Vatech

machine.

Angle number


Angle nomination Vatech Focal Trough Specific extrapolation

Vatech OPG extrapolation

Deference

1 17-16 Molar-Molar -9.2 4.8 -14.1

2 16-15 Premolar-Molar 30.8 -9.4 40.3

3 15-14 Premolar-Premolar -37.4 15.6 -53.0

4 14-13 Premolar-Anterior 11.8 -0.4 12.3


6 12-11 Anterior-Anterior 3.6 -2.1 5.8



9 22-23 Anterior-Anterior 1.5 2.3 -0.8


11 24-25 Premolar-Premolar -40.0 0.3 -40.3

12 25-26 Premolar-Molar 29.6 -1.5 31.2

13 26-27 Molar-Molar -16.9 -13.9 -3.0

14 37-36 Molar-Molar 3.3 3.7 -0.3

15 36-35 Premolar-Molar -11.2 -22.3 11.0





20 31-41 Anterior-Anterior -4.6 0.5 -5.1





25 45-46 Premolar-Molar -16.6 -13.2 -3.3

26 46-47 Molar-Molar -6.2 -6.9 0.6

61

Discussion

Discussion for Aim 1: Establish that CBCT is an effective way to assess tooth

angulations

The three CBCT and photo trial measurements were performed with at least an intervening day.

The measurement differences were not clinically significant which indicated a good intra-

examiner reliability level. There was no need to test the inter-examiner reliability as this was not

with-in the aims of this study. Comparing the values from Table 2 and Table 4, it was evident that

differences were not clinically significant. The differences ranged from 5 to -2 degrees with an

average difference (mean) of 0.6 degrees and a mode of 1 degree indicating that the CBCT was an

effective way to measure tooth angulations in space. The photographic measurement method was

taken as the standard for comparison, being the closest to reality after a Coordinate Measuring

Machine (CCM) which was unavailable.

Discussion for Aim 2: Define a method to establish the brand specific OPG focal

trough

As can be seen in Figure 29, and consistent with other reported findings [29, 30], the two different

panoramic machines (Carestream and Vatech) have two different focal troughs. They differed in

both shape and dimensions. Consequently their central planes varied in shape and dimension. The

Carestream focal trough looked closer to the shape of a dental arch with uniform trough thickness

and parallel distal ends while the Vatech focal trough had a rounder form with variable

thicknesses and converging distal ends.

The method established in this study is applicable to any type and brand of panoramic machine.

However, the focal trough was determined at the level of the occlusal plane and it is unknown

whether the focal trough is the same at different vertical levels. This requires further investigation.

Discussion for Aim 3: Generate a method to reveal how dentists mis-interpret the

OPG and to produce OPG and Focal Trough Specific angulations extrapolations.

Dentists consider OPGs as wrapped around the patient‟s head and then laid flat. The process of

making a panoramic radiograph involves much more.

In Figure 44, few dentists would appreciate that the Green arrows point to the medial side of the

mandibular condylar head (not the posterior side) and the blue arrows point at the lateral side of

the mandibular condylar head (not the anterior side)[69]. Also note how the same condylar head

appears different in two different panoramic images coming from two different panoramic

62

machines. It is important to appreciate how the panoramic images are created and the fact that

different panoramic machines have different focal troughs.

When interpreting panoramic images there is a missing link between how they are created and

their interpretation. It was, therefore, necessary to compare those two paradigms (conceptions).

The transfer of the coronal and apical tooth coordinates from their true 3D orientation into the 2D

representation “Focal Trough Specific angulation extrapolation” allows for the calculation of the

angles between adjacent teeth on that representation and to then compare them to their

correspondents on the “OPG extrapolations”. In other words, the OPG-extrapolated (reproduced)

image can be compared with the image produced after correction for the distortion introduced by

focal trough geometry.

63

Figure 44. Skull OPGs (A) Vatech machine (B) Carestream machine. Note that the Green arrows point at the medial side of the

mandibular condylar head and the blue arrows point at the lateral side of the mandibular condylar head. Also note how the

same condylar head look different in two different panoramic machine images.

64

Discussion for Aim 4: Compare the angulations extracted from the OPG


Visual comparison

For the visual comparison, this is the first time tooth OPG orientations have been visually

compared with their actual orientation. It is quickly appreciated how far panoramic presentation

differs from reality.

Angle measurements comparison

It may be appreciated that only the OPG extrapolation comparing the Carestream and Vatech

machines has all its values within a clinically acceptable limit of -5 to 5 degrees of angle

difference (Several authors reported that variations of 2.5° in either direction are clinically

acceptable [61, 70] but it was considered that this was too small to be clinically significant [71]).

Limits of agreement are provided in Table 9 for the 4 comparisons. It is obvious that the OPG

extrapolation comparison of Carestream versus Vatech has a much smaller limit of agreement

than the other 3 comparisons.

The level of agreement varied between the different tooth angle groups (Anterior-Anterior,

Anterior-Premolar, Premolar-Premolar, Premolar-Molar and Molar-Molar) when assessment

occurred in the Focal Trough Specific or OPG extrapolations when using the Carestream or

Vatech machines. It must be remembered that the Carestream Focal Trough Specific extrapolation

and Vatech Focal Trough Specific extrapolation were produced after applying the brand specific

OPG machine focal trough to the 3D (CBCT) scan coordinates.

OPG extrapolation Carestream versus OPG extrapolation Vatech

The different tooth-angle groups related well in this comparative study, both visually and

statistically. The results agreed with the findings of Mckee [46] who indicated that all OPG

machines were „wrong‟, but they are all wrong in the same way. All tooth group (Anterior-

Anterior, Anterior-Premolar, Premolar-Premolar, Premolar-Molar and Molar-Molar) relations

were consistent.

Focal Trough Specific extrapolation Carestream versus Focal Trough Specific extrapolation

Vatech

The variable examined was the shape of the focal trough. If it is considered that those

extrapolations were created to simulate how OPG images are interpreted by a dentist, it can be

appreciated how much diagnosis can be misinterpreted by looking at the same object from

different angles when following different focal troughs. The greatest discrepancy was in the

Premolar-Premolar group followed by the Anterior-Premolar group. All other group

65

measurements (Anterior-Anterior, Premolar-Molar and Molar-Molar) were consistent between

Carestream and Vatech Focal Trough Specific extrapolations.

Focal Trough Specific extrapolation versus OPG extrapolation for the Carestream machine

The comparison of the Focal Trough Specific and OPG extrapolation shows how OPGs can be

deceptive in indicating tooth relationships especially in the Premolar-Molar area and Molar-Molar

area. For example, in the OPG (Figure 34) teeth 35 and 36 appeared more parallel on the Focal

Trough Specific extrapolation than what they appeared on the OPG extrapolation. They were

divergent toward the apex on the OPG. The Premolar-Premolar, Anterior-Anterior and Premolar-

Molar angles related well. The probable reason for that is the shape of the focal trough and

whether the examined typodont fitted well inside the focal trough for the consistent regions but

was out for those which were inconsistent. This can be appreciated by looking at Figure 24 and

observing how some of the metal markers are clearer and defined than others. Further studies are

needed to measure the effect of different arch forms inside the same focal trough on the accuracy

of radiographic interpretation.

Focal Trough Specific extrapolation versus OPG extrapolation for the Vatech machine

By applying the same method to the Vatech machine, the areas of close relationship and poor

relationship were different. The Premolar-Premolar followed by Anterior-Premolar group did not

reveal high levels of agreement but Anterior-Anterior, Molar-Premolar and Molar-Molar groups

did show improved relationships. The reason can be explained by the paragraph above.

Discussion for Aim 5: Correct the OPG interpretation by forming an OPG brand-

specific formula.

Carestream

From the information displayed in Table 14, the following angle differences should be considered

when interpreting Carestream OPG images:

1- Molar-Molar relation: This relation could range from 7-16 degrees difference of

convergence toward the apices.

2- Premolar-Molar relation: This relation could range from 11-20 degrees difference of

divergence toward the apices.

3- Premolar-Premolar relation: This relation could range from 18 degrees difference of

convergence to 13 degrees difference of divergence towards the apices.

66

4- Premolar-Anterior relation: This relation could range from 16 degrees difference of

convergence to 4 degrees difference of divergence toward the apices.

5- Anterior-Anterior relation: This relation could range from 24 degrees difference of


Vatech

As can be seen in Table 15, the following angle differences should be considered when

interpreting Vatech OPG images:

1- Molar-Molar relation: This relation could range from 0-14 degrees difference of

convergence toward the apices.

2- Premolar-Molar relation: This relation could range from 3 degrees difference of

convergence to 40 degrees of divergence toward the apices.

3- Premolar-Premolar relation: This relation could range from 53 degrees difference of

convergence to 10 degrees difference of divergence towards the apices.

4- Premolar-Anterior relation: This relation could range from 5 degrees difference of


5- Anterior-Anterior relation: This relation could range from 19 degrees difference of


By applying the observed degrees of difference range, it is possible to correct OPG analysis to a

more realistic interpretation. The differences provide a much better understanding of the distortion

phenomenon when teeth appear to be clinically very well aligned but the OPG suggests that more

root or crown mesialization or distalization is required. This creates doubt and leaves clinicians

questioning whether the OPG indicators should be followed for treatment adjustments or

alternatively, the clinical appearance of the teeth could be relied upon and no adjustments

required. The present study has provided an answer by simply following OPG error range charts

that should be provided by the panoramic unit manufacturers and then an operator can make an

informed decision whether to adjust treatment or not.

The angulation differences may be affected by other factors including the patient‟s head position

[72], the radiographer‟s skill, dental arch dimensions and whether the focal trough is breeched.

Each brand of OPG unit has its own focal trough and hence produces images characteristic for

that machine. The manufacturers of panoramic units assume that the focal trough will match the

patients‟ dental arches[27]. Although average human dental arch measurements are used as a

reference, there is still a variable range. Nummikoski et al studied three different ethnic groups

(Mexican American, black American, and white American) and found the mean difference

between the groups ranged from 0.6 mm to 2.0 mm in dental arch widths. The standard deviation

was 1.5 mm within each study group. The male dentition was 0.6 mm wider in the cuspid and the

premolar areas compared with that of the female dentition, and the difference increased to 0.8 to

1.1 mm toward the posterior region of the arch. The different ethnic groups had a similar pattern

of width increase toward the posterior dentition but the dental arch width widens more in males.

Figure 45 illustrates their findings [73]. Welander et al recommended a mathematical expression

67

to describe the average form and size of the dentition and mandibular arc in Mexican American,

black American, and American and Scandinavian Caucasians that may be used by the panoramic

radiograph machine manufacturers. It was found that differences between the arch form means of

various ethnic groups where minimal and jaw size determined whether the dental arch would fit

within the focal trough [74]. In addition, the similarly manufacturer machines may produce

different central plane locations and even perhaps before and after calibration [75]. It is therefore

not possible to accept that all patients would fit and focus within a particular focal trough and

produce similar OPG images that would be realistically representative. Consequently variation in

arch form relative to the focal trough would produce differences in the apparent tooth angulations

seen on the OPG. This statement requires future studies.

Figure 45 Dental arches (D) and mandibular bone arches (M) plotted by gender (A) and by ethnicity (B) from Nummikoski et

al.[73]

68

Conclusion

The following conclusions were derived from the present study:

Conclusion for Aim 1: Establish that CBCT is a good way to assess teeth

angulations

CBCT is an acceptable way to estimate the angulation between different structures.

Conclusion for Aim 2: Define a method to establish the brand specific OPG focal

trough

Different OPG machines have different focal trough shapes and sizes.

Conclusion for Aim 3: Generate a method to reveal how dentists mis-interpret the

OPG and to produce OPG and Focal Trough Specific extrapolations.

It took slightly complex software computing to generate the “Focal Trough Specific

extrapolations”. Those extrapolations were clinically and statistically different to “OPG

extrapolations” leading to the conclusion in the next paragraph.

Conclusion for Aim 4 and Aim 5: Compare the angulations extracted from OPG

extrapolation to the Focal Trough Specific extrapolations and Correct the OPG

interpretation by forming an OPG brand-specific formula.

Panoramic radiographs were found to be not a reliable way to interpret the angulation between

teeth. The error range was:

1- For the Carestream machine: depending on the location within the arch from, divergence

ranged from 3 to 20 degrees and convergence ranged from 7 to 24 degrees.

2- For the Vatech machine: depending on the location within the arch from, divergence

ranged from 6 to 40 degrees and convergence ranged from zero to 53 degrees.

Manufacturers of OPG machines need to provide an error range for their devices that would

allow OPG interpreters better understand of the radiographic images.

Although there are guidelines supporting the safe use of CBCT in orthodontics as a useful method

of assessment [76, 77], there is still little justification [78-81] to send every patient for a CBCT

craniofacial image. Panoramic radiographs will remain in use for the foreseeable future but their

diagnostic interpretation should be approached with caution especially when assessing tooth

angulation.

This study added to our knowledge by indicating care when interpreting panoramic radiographs.

The image distortion is not only limited to the canines and premolars as was previously advised.

69

Depending on the panoramic machine and its focal trough, the range of error could be anywhere

in the dental arch and it should be the responsibility of the manufacturers to test their machines to

decrease the range of error or, at least, provide the users with that range.

More research is needed in (but not limited to) the following areas and their effect on tooth root

alignment in panoramic machines:

1- Patient‟s head position.

2- Different dental arch form shapes.

3- Different occlusal relationships (class I, class II or class III).

4- Whether the focal trough behaves the same at the occlusal plane level as well as at the

apical level.

5- Whether tooth arrangement (parallel versus arbitrary) affects angulation measurements.

70

References

1. Andrews, L.F., The six keys to normal occlusion. American journal of orthodontics, 1972. 62(3): p. 296-309.

2. Casko, J.S., et al., Objective grading system for dental casts and panoramic radiographs. American Journal of Orthodontics and Dentofacial Orthopedics, 1998. 114(5): p. 589-599.

3. Grading System for Dental Casts and Panoramic Radiographs June 2012, The American Board of Orthodontics.

4. Graber, T.M., Panoramic radiography in orthodontic diagnosis. American journal of orthodontics, 1967. 53(11): p. 799-821.

5. URSI, W.J.S., et al., Assessment of Mesiodistal Axial Inclination through Panoramic Radiography. JCO, 1990. 24(3): p. 166-173.

6. Burstone, C.J., Diagnosis and treatment planning ofpatients with asymmetries. Seminars in Orthodontics, 1998. 4(3): p. 153-164.

7. Farman, A.G., Tooth eruption and dental impactions. Panoramic Imaging News, 2004. 4(2): p. 1-7. 8. Farman, A.G., Panoramic radiologic appraisal of anomalies of dentition:Chapter #1. Panoramic

Imaging News, 2003. 3(1): p. 1-7. 9. Farman, A.G., Panoramic radiologic appraisal of anomalies of dentition:Chapter #2. Panoramic



Imaging News, 2004. 4(1): p. 1-7. 12. Franchi, L., Assessing growth and development with panoramic radiographs and cephalometric

attachments: a critical tool for dental diagnosis and treatment planning. Panoramic Imaging News, 2004. 4(4): p. 1-11.

13. Farman, A.G., Panoramic Radiology, Seminars on Maxillofacial Imaging and Interpretation2007, Heidelberg: Springer.

14. Paatero, Y.V., A New Tomographical Method for Radiographing Curved Outer Surfaces. Acta Radiologica [Old Series], 1949. 32(2-3): p. 177-184.

15. Blackman, S., Panoramic radiography. British Journal of Oral Surgery, 1963. 1(0): p. 209-218. 16. Hudson, D.C., J.W. Kumpula, and G. Dickson, A Panoramic X-ray Dental Machine. U.S. Armed

Forces Medical Journal., 1957. 8: p. 46-55. 17. Paatero, Y., Pantomography and orthopantomography. Oral Surgery, Oral Medicine, Oral

Pathology, 1961. 14(8): p. 947-953. 18. Stockley, S.m., A Manual of Radiographic Equipment 1986: Churchil Livingstone. 19. KOONG, B., Cone Beam Imaging in Dental Files2012, ADA Inc. 20. Scarfe, W.C., et al., Radiographic interproximal angulations: Implications for rotational panoramic

radiography. Oral Surgery, Oral Medicine, Oral Pathology, 1993. 76(5): p. 664-672. 21. Tronje G, W.U., McDavid WD, Morris CR., Imaging characteristics of seven panoramic x-ray units.

Projection angle. Dentomaxillofacial Radiology, 1985. 8(Suppl ch III): p. 21-28. 22. Lucchesi, M.V., R.E. Wood, and C.J. Nortjé, Suitability of the panoramic radiograph for assessment

of mesiodistal angulation of teeth in the buccal segments of the mandible. American Journal of Orthodontics and Dentofacial Orthopedics, 1988. 94(4): p. 303-310.

23. Garcia, M., The effect of buccolingual root angulation on the mesiodistal image perception for panoramic images. American Journal of Orthodontics and Dentofacial Orthopedics, 2005. 128(2): p. 262.

24. Garcia-Figueroa, M.A., et al., Effect of buccolingual root angulation on the mesiodistal angulation shown on panoramic radiographs. American Journal of Orthodontics and Dentofacial Orthopedics, 2008. 134(1): p. 93-99.

71

25. Garcia-Figueroa, M.A., et al., Measurement of mesiodistal root angulation for panoramic images and the effect of buccolingual root angulation. International orthodontics / College europeen d'orthodontie, 2009. 7(1): p. 15-30.

26. Lee, J., The effects of buccolingual root torque on the appearance of root angulation on panoramic radiographs. American Journal of Orthodontics and Dentofacial Orthopedics, 2005. 127(3): p. 393.

27. Glass, B.J., et al., The central plane of the image layer determined experimentally in various rotational panoramic x-ray machines. Oral Surgery, Oral Medicine, Oral Pathology, 1985. 60(1): p. 104-112.

28. Scarfe, W.C., F.E. Eraso, and A.G. Farman, Characteristics of the Orthopantomograph OP 100. Dentomaxillofacial Radiology, 1998. 27(1): p. 51-7.

29. Lund, T.M. and L.R. Manson-Hing, A study of the focal troughs of three panoramic dental x-ray machines: Part I. The area of sharpness. Oral Surgery, Oral Medicine, Oral Pathology, 1975. 39(2): p. 318-328.

30. Lund, T.M. and L.R. Manson-Hing, A study of the focal troughs of three panoramic dental x-ray machines: Part II. Image dimensions. Oral Surgery, Oral Medicine, Oral Pathology, 1975. 39(4): p. 647-653.

31. Rejebian, G.P., A statistical correlation of individual tooth size distortions on the orthopantomographic radiograph. American journal of orthodontics, 1979. 75(5): p. 525-534.

32. Hellman, M., Dimension s. Form in Teeth and Their Bearing on the morphology of the Detal Arch. International Journal of Orthodontics, 1919. 7: p. 615-651.

33. Meredith, H.V. and L.B. Higley, Relationships Between Dental Arch Width and Widths of the Face and Head, . American Journal of Orthodontics, 1951. 37: p. 193-204.

34. Bonwill, W.G.A., Geometrical and Mechanical Laws of Articulation Transactions of the Odontological Society of Pennsylvania, 1884-1885: p. 119-133.

35. Hawley, C.A., Determination of the Normal Arch and Its Application to Orthodontia. Dental Cosmos, 1905. 47: p. 541-552.

36. Williams, P.M., Determming the Shape of the Normal Arch. Dental Cosmos, 1917. 59: p. 695-708. 37. Lasher, M.C., A Consideration of the Principles of Mechanical Arches as Applied to the Dental Arch.

Angle Orthodontist, 1934. 4: p. 248-268. 38. Paatero, Y.V., Pantomography in Theory and Use. Acta Radiol, 1954. 41: p. 331-335. 39. Bromwell, I.N., Anatomy and Histology of the Mouth and Teeth. 2 ed1902, Philadelphia: P.

Blakiston’s Son & Comnany. . 40. Angle, E.H., Treatment of Malocclusion of the Teeth. 7 ed1907, Philadelphia: S.S. White Dental

Mfg. Co. 41. Wheeler, R.C., A Textbook of Dental Anatomy and Physiology. 2 ed1950, Philadelphia: W. B.

Saunders Company. 42. Black, G.V., Descriptive Anatomy of the Human Teeth. . 5 ed1902, Philadelphia: White Dental

Mfg. Co. . 43. Izard, G., New Method for Determination of the Normal Arch by the Function of the Face. .

International Journal of Orthodontics, 1927. 13: p. 582-595. 44. Lund, T.M. and L.R. Manson-Hing, Relations between tooth positions and focal troughs of

panoramic machines. Oral Surgery, Oral Medicine, Oral Pathology, 1975. 40(2): p. 285-293. 45. Manson-Hing, L.R., T.M. Lund, and T. Ohba, Japanese tooth positions and their relation to

panoramic radiography. Oral Surgery, Oral Medicine, Oral Pathology, 1976. 41(6): p. 797-802. 46. McKee, I.W., et al., The accuracy of 4 panoramic units in the projection of mesiodistal tooth

angulations. American Journal of Orthodontics and Dentofacial Orthopedics, 2002. 121(2): p. 166-175.

47. McDavid, W.D., Imaging characteristics of seven panoramic X-ray units1985: s.n. 48. Philipp, R.G. and R.V.V. Hurst, The Cant of the Occlusal Plane and Distortion in the Panoramic

Radiograph. The Angle orthodontist, 1978. 48(4): p. 317-323.

72

49. Van Elslande, D., et al., Accuracy of mesiodistal root angulation projected by cone-beam computed tomographic panoramic-like images. American Journal of Orthodontics and Dentofacial Orthopedics, 2010. 137(4, Supplement): p. S94-S99.

50. Tong, H., et al., A new method to measure mesiodistal angulation and faciolingual inclination of each whole tooth with volumetric cone-beam computed tomography images. American Journal of Orthodontics and Dentofacial Orthopedics, 2012. 142(1): p. 133-143.

51. Tong, H., et al., Mesiodistal angulation and faciolingual inclination of each whole tooth in 3-dimensional space in patients with near-normal occlusion. American Journal of Orthodontics and Dentofacial Orthopedics, 2012. 141(5): p. 604-617.

52. DeHaan, D.A., R. Kulbersh, and R. Al-Qawasmi, A New cone -beam computed tomography technique to evaluate full -tooth angulation and inclination in orthodontic cases, 2013, Nova Science Publishers, Inc. p. 343-355.

53. Panmekiate, S., W. Apinhasmit, and A. Petersson, Effect of electric potential and current on mandibular linear measurements in cone beam CT. Dentomaxillofacial Radiology, 2012. 41(7): p. 578-582.

54. Ostermiller, S. Triangle Calculator. ostermiller 2006; Available from: http://ostermiller.org/calc/triangle.html.

55. Wise, S.P., R. Kulbersh, and R. Kaczynski, Mesiodistal root angulation shown on conventional and reconstructed panoramic radiographs, in Computed Tomography: New Research2013, Nova Science Publishers, Inc. p. 325-342.

56. Pittayapat, P., et al., Agreement between cone beam computed tomography images and panoramic radiographs for initial orthodontic evaluation. Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology, 2014. 117(1): p. 111-119.

57. Bouwens, D.G., et al., Comparison of mesiodistal root angulation with posttreatment panoramic radiographs and cone-beam computed tomography. American Journal of Orthodontics and Dentofacial Orthopedics, 2011. 139(1): p. 126-132.

58. Leuzinger, M., et al., Root-contact evaluation by panoramic radiography and cone-beam computed tomography of super-high resolution. American Journal of Orthodontics and Dentofacial Orthopedics, 2010. 137(3): p. 389-392.

59. contributors, W. 29 January 2014; Available from: http://en.wikipedia.org/wiki/Cobb_angle. 60. JR., C., Outline for the study of scoliosis. The American Academy of Orthopedic Surgeons

Instructional Course Lectures. Vol. 5. 1948: Ann Arbor. 61. McKee, I.W., The accuracy of panoramic radiography in the assessment of mesiodistal tooth

angulations at varying horizontal and vertical head positions, 2001, National Library of Canada = Bibliothèque nationale du Canada.

62. Owens, A.M. and A. Johal, Near-End of Treatment Panoramic Radiograph in the Assessment of Mesiodistal Root Angulation. The Angle orthodontist, 2008. 78(3): p. 475-481.

63. Stramotas, S., et al., Accuracy of linear and angular measurements on panoramic radiographs taken at various positions in vitro. The European Journal of Orthodontics, 2002. 24(1): p. 43-52.

64. Rasband, W.S., ImageJ, 1997-2012, U. S. National Institutes of Health: Bethesda, Maryland,USA. 65. Eric Jones, T.O., Pearu Peterson and others., SciPy: Open Source Scientific Tools for Python, 2001. 66. Altman, D.G. and J.M. Bland, Measurement in Medicine: The Analysis of Method Comparison

Studies. Journal of the Royal Statistical Society. Series D (The Statistician), 1983. 32(3): p. 307-317. 67. Bland, M.J. and D. Altman, STATISTICAL METHODS FOR ASSESSING AGREEMENT BETWEEN TWO

METHODS OF CLINICAL MEASUREMENT. The Lancet, 1986. 327(8476): p. 307-310. 68. Bland, J.M. and D.G. Altman, Statistical methods for assessing agreement between two methods

of clinical measurement. International Journal of Nursing Studies, 2010. 47(8): p. 931-936. 69. Monsour, P. Cone beam (CB) CT radiography. Getting the most out of the image. in Twilight

Seminar series. 2014. Dandenong. Victoria: A. Professor Phillippe Zimet. 70. Tronje, G., et al., Image distortion in rotational panoramic radiography. III. Inclined objects. Acta

Radiol Diagn (Stockh), 1981. 22(5): p. 585-92.

http://ostermiller.org/calc/triangle.html

http://en.wikipedia.org/wiki/Cobb_angle

73

71. Welander, U., et al., Inclined Objects, in Imaging Characteristics of Seven Panoramic X-ray Units, W.D. McDavid, et al., Editors. 1985, Dento-maxillo-facial radiology: Umeå,Sweden. p. 45-50.

72. McKee, I.W., et al., The Effect of Vertical and Horizontal Head Positioning in Panoramic Radiography on Mesiodistal Tooth Angulations. Angle Orthodontist, 2001. 71(6): p. 442-451.

73. Nummikoski, P., et al., Dental and mandibular arch widths in three ethnic groups in Texas: A radiographic study. Oral Surgery, Oral Medicine, Oral Pathology, 1988. 65(5): p. 609-617.

74. Welander, U., et al., Standard forms of dentition and mandible for applications in rotational panoramic radiography. Dentomaxillofacial Radiology, 1989. 18(2): p. 60-7.

75. Razmus, T.F., B.J. Glass, and W.D. McDavid, Comparison of image layer location among panoramic machines of the same manufacturer. Oral Surgery, Oral Medicine, Oral Pathology, 1989. 67(1): p. 102-108.

76. Clinical recommendations regarding use of cone beam computed tomography in orthodontics. [corrected]. Position statement by the American Academy of Oral and Maxillofacial Radiology. Oral Surg Oral Med Oral Pathol Oral Radiol, 2013. 116(2): p. 238-57.

77. Scarfe, W.C., Clinical recommendations regarding use of cone beam computed tomography in orthodontic treatment. Position statement by the American Academy of Oral and Maxillofacial Radiology. Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology, 2013. 116(2): p. 238-257.

78. CONE BEAM CT FOR DENTAL AND MAXILLOFACIAL RADIOLOGY, 2012, EUROPEAN COMMISSION Directorate General for Energy Directorate D - Nuclear Energy , Unit D4 - Radiation Protection Luxembourg.

79. Hodges, R.J., K.A. Atchison, and S.C. White, Impact of cone-beam computed tomography on orthodontic diagnosis and treatment planning. American Journal of Orthodontics and Dentofacial Orthopedics, 2013. 143(5): p. 665-674.

80. Horner, K., et al., Basic principles for use of dental cone beam computed tomography: consensus guidelines of the European Academy of Dental and Maxillofacial Radiology. Dentomaxillofac Radiol, 2009. 38(4): p. 187-95.

81. Silva, M.A.G., et al., Cone-beam computed tomography for routine orthodontic treatment planning: A radiation dose evaluation. American Journal of Orthodontics and Dentofacial Orthopedics, 2008. 133(5): p. 640.e1-640.e5.

Documents

Evaluation of tooth angulation measured on cone beam … · 2014-10-16 · Evaluation of tooth angulation measured on cone beam computed tomography and panoramic radiographs A thesis